VARIATION IN THE DESMIDIE X. 367 
of the mother and of a newly developed half. The latter is 
sometimes markedly different from the parent half, but in cases 
where this difference is extreme the new halves of the next 
division generally conform to the original type, though more 
rarely this may not occur for several generations, 
The great diversity of form and wonderfuily varied character 
of these plants are to be associated with their confinement to 
small ponds or the quiet margins of lakes, &c., localities suitable 
for their existence in large numbers. In these restricted areas 
the unceasing effect of the struggle for existence will result 
in a gradually increasing diversity of form, and this is to be 
correlated with the immense numbers of individuals that are 
sometimes found in these situations. 
There can be little doubt that at the present day the family 
of Conjugates which has attained a maximum state of specializa- 
tion along one line, accompanied by degeneration along another, 
is the Desmidiew ; and also that the tendency has been in the 
direction of an increase in the complexity of morphological 
characters. May not this complexity of outline, which is so fre- 
quently accompanied by a complete defensive armour of spines 
and spinous processes, have been acquired as a means of resisting 
the attacks of many forms of aquatic animals? After the loss 
of the filamentous condition it became necessary that the solitary 
and unprotected individuals should aequire some means of 
defence. Presumably many forms developed spines and warts, 
or their outline became deeply incised, and, by a gradual process 
of natural selection or survival of the fittest, the present mor- 
phological complexity was in all probability for the most part 
brought about. These plants are devoured by many small aquatic 
animals—Ameebe, Turbellaria, Oligochetes, Tardigrades, Crus- 
tacea, &c.; and it is a notable fact that those species occurring 
on wet rocks and other localities in which these enemies are 
either absent or very scanty, especially at high elevations, possess 
as a rule a comparatively simple outline. In such cases the 
plants are provided with a more or less abundant mucus, whereas 
especially prior to conjugation. Another fact testifying to this degeneration 
is the secondary assumption of the filamentous condition by certain species 
(e. g., Micrasterias foliacea) and the reversion in others (e.g., Hyalotheca 
dissiliens) to a type ef conjugation with sexually differentiated gametes. 
